Leitech AI Hardware Group | Editor: Dingxi | Supervisor: Luo Chao
The "range anxiety" of drones finally has a new solution in 2026.
Over the past few years, drones have been increasingly applied in urban governance, forest fire prevention, power line inspection, and other fields. Especially with the popularization of unattended docks, drones can now take off, execute missions, and return on their own, even completing data analysis automatically. However, one issue has never been thoroughly resolved—endurance.
No matter how strong the flight performance of a drone is, it must return for charging when the battery is depleted. For ordinary aerial photography users, waiting for half an hour might not matter much, but in scenarios requiring continuous monitoring, such as forest fire prevention and emergency rescue, this half-hour gap could cause significant impacts.
On October 14, DJI officially released the new generation unattended platform—DJI Dock 4. This time, DJI brought a quite interesting technology: 25-second automatic battery swap. In addition, the new generation Matrice 5TD drone has also seen comprehensive upgrades in endurance, imaging, environmental perception, and AI capabilities. On the software side, FlightHub 2 provides operation management tools, while the brand-new Cloud API 2 is aimed at developers, helping them integrate drone operations into different industry systems. DJI hopes to enable drones to take on more daily tasks and even become a standard configuration in urban management.
On the eve of the new product launch, Leitech (ID: Leitech) was also invited to Wutong Mountain in Shenzhen to get a close-up look at the various new technologies of DJI Dock 4 and participate in the on-site experience event. So, what exactly is worth paying attention to in this new generation unattended platform? And what changes can it bring to the rapidly developing low-altitude economy?
25-Second Automatic Battery Swap, DJI Solves the Drone Endurance Problem
Readers familiar with DJI enterprise drones should know that the DJI Dock is not a new product. As early as 2022, DJI launched the first-generation drone dock, enabling drones to automatically complete a series of tasks such as takeoff, cruising, returning, and charging without on-site pilot operation. By the time of DJI Dock 3 in 2025, drones had already been able to achieve quite mature automated operations.
It can be said that DJI's solution is already quite mature and at a leading level in the industry. But it still has one problem: the charging time is too long. According to data released by DJI, the previous generation Dock 3 takes about 30 minutes to complete a single charge. Some readers might think that half an hour is not that long, right? But it's a different story in another scenario. Suppose a drone has just completed a forest patrol, and a suspected fire suddenly breaks out nearby, but the drone is still charging in the dock. That half-hour could delay a lot of things. For forest fire prevention, traffic management, and emergency rescue, whether the drone can be dispatched at the very first moment is even more important than its flight speed.
DJI's solution this time is very simple: automatic battery swap.
At the technology sharing event at Wutong Mountain in Shenzhen, DJI also arranged an on-site demonstration of the Dock 4's automatic battery swap function. According to the on-site staff, DJI Dock 4 adopts a turntable-style battery swap structure. After the drone lands, the dock will automatically remove the old battery and then install the fully charged battery in place. The entire process takes only 25 seconds. What's even more interesting is that DJI has also added a wireless power supply design. Even if the battery has been removed, the drone's avionics system can still keep running, saving the time for rebooting, and it can take off again in as fast as 45 seconds.
For Xiao Lei, who is often in contact with consumer drones, manually replacing the battery is certainly not a difficult task. But if the drone needs to be deployed in mountains, power stations, and other places for a long time, specially arranging staff to go there and swap batteries would obviously be quite troublesome.
Of course, fast battery swapping alone is not enough; the drone's own endurance cannot be too poor. The matched Matrice 5TD this time has a maximum endurance of 59 minutes and an operating radius of 10 kilometers. DJI also stated that the theoretical operating duty cycle of the new generation dock can reach 98.4%. Coupled with relay scheduling between multiple docks, the monitoring gap can be further reduced.
In the past, drones had to go back and rest after flying for a certain period. Now, it only takes tens of seconds to get back to work. If a city deploys enough docks and several drones take turns going out, continuous aerial monitoring of key areas can be achieved.
Clear License Plates at 400 Meters, Major Upgrade to the "Eyes" of DJI Drones
Apart from the dock itself, the Matrice 5TD, which made its debut simultaneously, is also worth talking about.
Unlike ordinary aerial photography drones, enterprise drones often face much more complex tasks. For example, checking whether the bolts on high-voltage towers are loose, finding abnormal heat sources hidden in the woods, or searching for missing persons at night. To complete these tasks, the drone first needs to see far enough and clearly enough, and preferably work normally at night.
To this end, DJI has equipped the Matrice 5TD with multiple modules such as telephoto, infrared thermal imaging, and laser rangefinding. Among them, the 11x telephoto lens can identify license plates at a distance of 400 meters, and can even detect small defects such as bolts and pins at a distance of over 100 meters. Everyone might not have a clear concept of this distance. Simply put, some details that previously required staff to climb the tower for inspection can now be clearly captured by the drone from a distance and then handed over to backend personnel for judgment. Especially for those transmission lines located in mountains and deserts, this can save a lot of time and manpower.
Night capability is also a key focus of this upgrade. The Matrice 5TD supports full-color 4K high-frame-rate night vision, and also has the dual-light fusion capability of infrared thermal imaging and visible light, which can combine the target outline and temperature information. Coupled with the laser rangefinding function with a maximum range of 3 kilometers, after discovering anomalies, staff can quickly mark the target's position and distance.
Take the event venue, Wutong Mountain, as an example. Its vegetation coverage is quite dense, and the mountain roads are relatively complex. If relying entirely on manual patrols, it is not easy to detect every anomaly in a timely manner. Suppose a suspected fire breaks out in a certain forest area. The drone can first use thermal imaging to find abnormal heat sources, then use the telephoto lens to observe the surrounding situation, and finally feed back the specific location to the ground staff. Similar functions can also be used in scenarios such as tourist search and rescue, and mountain forest patrols. Especially for some locations that are difficult for ground personnel to reach in a short time, the advantages of drones will be even more obvious.
In addition, the Matrice5TD also supports up to three payload mounts, which can be paired with devices such as loudspeakers and searchlights. For example, when searching for missing tourists at night, the drone can not only find the target through thermal imaging but also use the searchlight for illumination and communicate with ground personnel through the loudspeaker. To ensure flight safety in complex environments, DJI has also added millimeter-wave radar, LiDAR, and a six-directional visual perception system. The dock itself supports an operating environment from -40°C to 50°C.
Apart from battery swapping and imaging, DJI has also put a lot of effort into AI this time.
Although drones in the past were already capable of automatic patrols, the captured footage still needed to be reviewed by staff. Especially in scenarios like urban management and forest fire prevention, a single drone might generate a large amount of photos and videos every day. If all of them rely on manual inspection, the workload is still considerable.
This time, DJI has added onboard AI recognition capabilities to the Matrice 5TD, which can identify 17 types of targets such as vehicles, ships, and smoke/fire during flight, and supports third-party custom models. In addition, the DJI FlightHub 2 platform has also introduced a multimodal large model. Staff can directly describe the targets they need to find through text or images, allowing the system to assist in recognition and alerting.
To give an example, in the past, when patrolling the forest for abnormal smoke points, staff might have to stare at the screen for a long time. Now, AI can be used to filter out suspicious footage first, and then staff can confirm it. For long-term patrol tasks, the system can also compare images taken at different times, find out the locations where changes have occurred, and generate corresponding reports.
At the technology sharing event at Wutong Mountain, DJI also highlighted the application directions of these AI capabilities. Xiao Lei believes that these functions are more meaningful for unattended docks than simply improving the flight performance of drones. After all, if a city deploys hundreds of docks and each drone requires a person to stare at the footage, the so-called unattended operation would obviously be a misnomer.
In fact, similar applications have already been implemented in Wuhan. Previously, Leitech learned in Wuhan that the local drone dock network has been providing services for 16 departments. Over the past year, it has participated in handling a total of 400 elevated highway traffic accidents, of which about 30% were guided to be handled by the car owners themselves through remote loudspeaker broadcasting. The drone is responsible for quickly arriving at the scene and transmitting the footage back, while the staff handles the situation according to the actual conditions, which indeed saves a lot of manpower.
Of course, at this stage, AI cannot completely replace manual judgment, especially when it involves public safety and traffic law enforcement, and the recognition results still need to be verified. But at least for those highly repetitive patrol tasks, AI has already been able to help staff reduce a lot of burden.
DJI is Making Drone Docks a New Type of Infrastructure
Looking back at the development of DJI Dock over the past few years, we can actually find a very obvious change. The first-generation DJI Dock solved the problem of whether drones can take off and return automatically; Dock 3 further improved automation capabilities and deployment flexibility; and now with Dock 4, DJI has begun to focus on solving issues such as continuous operations, rapid response, and AI management.
According to data released by DJI, the DJI Dock series has currently entered more than 80 countries and regions globally, completing a cumulative total of over 8 million flight sorties. At least from the data, drone docks have gradually stepped out of the early technology validation stage and begun to take on more and more practical work.
But deploying many docks does not mean that various departments can use them directly. The transportation department needs to integrate drones into the accident handling process, the forestry department focuses on fire alerts, and power enterprises have their own inspection and defect management systems. The same drone, when entering different businesses, requires different software and processing processes behind it. This is also the reason why DJI launched Cloud API 2 this time.
Simply put, it is aimed not only at ordinary pilots but also at the development teams of enterprise customers, software companies, and integrators. According to DJI, this set of development tools reduces repetitive work in device access and basic function construction through AI-assisted programming, reusable general modules, and front-end components such as dashboards and route editors, allowing developers to put more energy into specific industry needs. For Xiao Lei, this matter is not as intuitive as the 25-second battery swap, but it is equally related to whether the dock can be popularized quickly: DJI does not need to develop complete software for every industry, and partners do not need to build similar basic functions from scratch in every project.
From urban governance in Wuhan to the on-site experience at Wutong Mountain this time, Xiao Lei's biggest feeling is that the uses of drones in the industry are indeed becoming more and more diverse. In the past, everyone cared about how far the drone could fly and how clearly it could shoot. Now, they also have to consider whether it can execute tasks stably for a long time. The 25-second automatic battery swap of DJI Dock 4 at least allows drones to no longer have to go back and charge for half an hour every now and then. This innovation by DJI is solid and represents a huge progress for the entire drone industry.